The Effect of Patient-Centered, Smartphone-Based Diabetes Care System

Abstract

The prevalence of diabetes mellitus is 8.3% worldwide, and the number of patients is continuously increasing. Medical costs related to diabetes, largely due to the treatment costs of diabetic complications resulting from poor glycemic control, represent a significant socioeconomic burden. Diabetes is a chronic disease that requires life-long management

therefore, it is important that strategies focus on the education of patients towards self-management and the support of daily activities. Various information technology (IT)-based interventions have been developed for chronic diseases including diabetes. However, the efficacy and safety of these interventions have not been established through well-designed randomized controlled clinical trials. Furthermore, current IT-based systems for disease management have limited functionalities for diabetes self-management. This study aimed to develop a multifunctional diabetes management system and assess its efficacy and safety. The patient-centered smartphone-based diabetes care system (PSDCS) contained 4 modules on glucose monitoring, diet, physical activity, and social network service. Each module collected data via a Bluetooth glucometer and activity tracker and gave appropriate feedback or a reminder according to preset algorithms. The system also incorporated a web site allowing medical staff to monitor individual input data. A single-arm pilot study aimed to test the feasibility of HbA1c reduction with the PSDCS over 12 weeks. Twenty-nine patients completed the pilot study. The HbA1c and fasting plasma glucose levels decreased significantly from baseline (7.7 ± 0.7% to 7.1 ± 0.6%, p < 0.0001

140.9 ± 39.1 mg/dL to 120.1 ± 31.0 mg/dL, p = 0.0088, respectively) after 12 weeks of intervention. The number of glucose measurement correlated with the extent of reduction in HbA1c levels (r = －0.53, p = 0.0013). The scores of the Summary of diabetes self-care activities including diet, exercise, and glucose monitoring, showed a significant improvement, particularly in the patients in the upper tertile of HbA1c reduction. No severe adverse event occurred during the intervention. Based on the results of the pilot study, we upgraded the system and conducted to a multicenter, randomized controlled study over 24 weeks. The upgraded system had an insulin dosing algorithm for basal insulin and premixed insulin, a rescue therapy protocol, and an improved user interface. A total of 172 patients, excluding the patients who missed the primary endpoint or did not fulfill the inclusion/exclusion criteria, were included in the analysis (full analysis set). After 24 weeks, HbA1c reduction relative to baseline was greater in the mHealth group (－0.4 ± 0.1%) compared to the standard care group (－0.1 ± 0.1%). The difference between the adjusted mean changes was 0.35% (95% CI: 0.14-0.55, p = 0.001). The reduction in HbA1c levels was more evident among patients with a baseline HbA1c of 8.0% or higher (－0.9 ± 0.2% vs. －0.3 ± 0.2%, p = 0.016) and who used insulin (－0.7 ± 0.2% vs. －0.2 ± 0.2%, p = 0.014). The proportion of patients achieving HbA1c levels below 7.0% was 41.1% for the mHealth group and 20.7% for the standard care group (OR = 2.01, 95% CI: 1.24-3.25, p = 0.003). The proportion of the patients achieving HbA1c levels of 6.5% or lower was 14.4% and 2.4%, respectively (OR = 5.78, 95% CI: 1.40-23.86, p = 0.004). Overall, 31.1% of patients in the mHealth group and 17.1% of patients in the standard care group achieved HbA1c levels below 7.0% without hypoglycemia (OR = 1.82, 95% CI: 1.03-3.21, p = 0.024). Among insulin users, the range of glucose fluctuation in the mHealth group was significantly reduced compared to that in the standard care group (－50.8 ± 60.6 mg/dL vs. 4.4 ± 62.0 mg/dL, p = 0.017). There was no difference in numbers of hyper- and hypoglycemic events between the groups. No patient was treated with rescue therapy, and system-related adverse events did not occur during the study. In conclusion, our smartphone-based PSDCS improved glucose control with a tolerable safety profile, and can be applied to patients with type 2 diabetes.